Extravasation of Intravenous Computed Tomography Scan Contrast in Blunt Abdominal and Pelvic Trauma Ivan R. Diamond, MD, Paul A. Hamilton, MD, Adam B. Garber, Homer C. Tien, MD, Talat Chughtai, MD, Sandro B. Rizoli, MD, PhD, Lorraine N. Tremblay, MD, PhD, and Frederick D. Brenneman, MD Background: Intravenous contrast extravasation (CE) on computed tomogra- phy (CT) scan in blunt abdominal trauma is generally regarded as an indication for the need for invasive intervention (either angiography or laparotomy). More recently, improvements in CT scan technology have increased the sensitivity in detecting CE, and, thus, we postulate that not all patients with this finding require intervention. Methods: This study is a retrospec- tive review of all patients who underwent a CT scan for blunt abdominal trauma between January 1999 and September 2003. Patterns of injury, associated inju- ries, management, and outcomes were ex- amined for patients with CE. Results: Seventy of 1,435 patients (4.8%) demonstrated CE. Mean age was 44 years and mean Injury Severity Score was 39. The location of CE was intra- abdominal in 25, pelvis/retroperitoneum in 39, and both areas in 3 patients. Six patients received supportive treatment for nonsurvivable head injury and were excluded from further analysis. Overall, 30 (47%) patients underwent immediate intervention (angiography or laparot- omy) and 34 (53%) were managed non- operatively. Of those who had initial nonoperative management, overall seven (20.5%) underwent intervention, with the remainder being managed without inter- vention. The success for nonoperative management was greater for those with pelvic/retroperitoneal CE (4 of 7: 57%) than for intra-abdominal extravasation (23 of 27: 85%). Conclusion: Although evidence of CE may suggest significant vascular injury, our data suggest that not all pa- tients require invasive intervention. Fur- ther studies are needed to better define criteria for nonoperative management in patients with CE identified on their initial CT scan. Key Words: Abdominal injuries, Pel- vic injuries, Extravasation of diagnostic and therapeutic materials, Computed tomography. J Trauma. 2009;66:1102–1107. H emodynamically stable patients with blunt abdomino- pelvic trauma routinely undergo computerized tomog- raphy (CT) scanning with intravenous and oral contrast as part of their initial assessment. 1,2 Intravenous contrast extravasation (CE), which presents as an area of diffuse or focal high density, isodense as compared with major adjacent vascular structures, represents active arterial or venous bleeding. 3 Generally, the finding of CE is viewed as an indication for invasive intervention— either laparotomy or angiography. 4 –11 We have recently observed a significantly higher inci- dence of CE than the 0.2% predicted from a large population- based study. 8 This may be related to advances in imaging technology, 3,4 or because of an increased awareness of this finding. However, given the increased incidence, we wished to examine the patterns of initial management and outcome of the patients at our center who had intravenous CE demon- strated on their initial trauma CT scan. We hypothesized that, given the increased detection, a significant number of pa- tients, despite having CE identified, could be managed with- out intervention. METHOD Patients We retrospectively reviewed the reports of all CT scans done for blunt abdominal trauma between January 1999 and September 2003. The decision to proceed to CT scan was at the discretion of the attending trauma team leader, with no particular protocol in effect. However, in general only hemo- dynamically stable patients, with suspected abdomino-pelvic injury, are subjected to CT scanning. All CT scans were performed using a General Electric 4 Multidetector CT scan- ner throughout the study period. Those patients with intrave- nous CE were identified for further review. Charts were reviewed to determine the presentation, initial management, and outcome of these patients. Patients were classified as operative if after the CT scan they underwent intervention, either laparotomy or angiogra- phy. All other patients were deemed to have a nonoperative strategy. This was further subclassified as successful or un- successful depending on whether there was a subsequent need to proceed to intervention in the 48 hours after initial trauma resuscitation. Initial management as decided upon by the attending trauma surgeon was documented. Submitted for publication November 14, 2007. Accepted for publication March 17, 2008. Copyright © 2009 by Lippincott Williams & Wilkins From the Departments of Surgery and Medical Imaging and the Trauma, Emergency and Critical Care Program, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada. Presented at the Annual Scientific meeting of the Trauma Association of Canada, April 6 –9, 2005 Whistler, British Columbia, Canada. Address for reprints: Frederick D. Brenneman, MD, Sunnybrook Health Sciences Centre, Room H-170, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5; email: fred.brenneman@sunnybrook.ca. DOI: 10.1097/TA.0b013e318174f13d The Journal of TRAUMA  Injury, Infection, and Critical Care 1102 April 2009